Endocrine Abstracts

Introduction: Presence of thyroid peroxidase antibodies (TPOAbs) may be a marker of future thyroid failure. However, prospective studies in a euthyroid population have not been performed.

Aim of the study: To prospectively investigate the relationship between presence and levels of TPOAbs and incident hypothyroidism in euthyroid subjects in the general population.

Study population: The database used for this study consisted of a random sample of 2703 participants of the PREVEND (prevention of renal and vascular end stage disease) study, inhabitants aged 28–75 years of the city of Groningen, The Netherlands. We excluded subjects with a TSH level outside the laboratory’s reference range (0.35–4.94 mIU/l; n=115) at baseline and subjects taking thyroid medication (n=37) and/or medications that may affect thyroid function (n=56). Incident hypothyroidism was defined as initiation of L-thyroxine therapy in the absence of thyreostatic medication by the participants during follow-up.

Results: Prevalence of positive TPOAbs at baseline was 8.5%, with levels between 12 and 3767 IU/ml. Prevalence significantly increased with increasing TSH concentrations (P<0.001). Median age at baseline was 46 years. Median follow up was 6.7 years. Incidence of hypothyroidism was 0.5% (3.0% in TPOAbs positive versus 0.2% in TPOAbs negative subjects; P<0.001). Sex (HR 8.6 (1.10–67.17), P=0.04), TSH level (HR 3.4 (2.07–5.60), P<0.001) and TPOAbs level (log transformed; HR 3.93 (2.26–6.82), P<0.001) were significant predictors of incident hypothyroidism in univariate analysis. Age was not related. When tested multivariately, the product term of TSH level and TPOAbs level was significantly related to incident hypothyroid cases (P=0.01), reflecting highest risk in subjects with both high levels of TPOAbs and high levels of TSH within the reference range.

Conclusion: We have demonstrated that TPOAbs level and TSH level are interdependent predictors for future hypothyroidism, even when TSH is still within the laboratory’s reference range.